A hierarchical clustered BitTorrent （CBT） system is proposed to improve the file sharing perior-mance of the BitTorrent system, in which peers are grouped into clusters in a large-scale BitTorrent-hke underlying ove...A hierarchical clustered BitTorrent （CBT） system is proposed to improve the file sharing perior-mance of the BitTorrent system, in which peers are grouped into clusters in a large-scale BitTorrent-hke underlying overlay network in such a way that clusters are evenly distributed and that the peers within the cluster are relatively close to each other. A fluid model is developed to compare the performance of the proposed CBT system with the BitTorrent system, and the result shows that the CBT system can effectively improve the performance of the system. Simulation results also demonstrate that the CBT system improves the system scalabihty and efficiency while retaining the robustness and incentives of the original BitTorrent paradigm.展开更多
In the manufacturing grid＇s architecture, Resources Management System （RMS） is the central component responsible for disseminating resource information across the grid, accepting requests for resources, discovering...In the manufacturing grid＇s architecture, Resources Management System （RMS） is the central component responsible for disseminating resource information across the grid, accepting requests for resources, discovering and scheduling the suitable resources that match the requests for the global grid resource, and executing the requests on scheduled resources. In order to resolve the problem of resources publication and discovery in Manufacturing Grid （MGrid）, the classification of manufacturing resources is first researched after which the resources encapsulation class modes are put forward. Then, a scalable two-level resource management architecture is constructed on the model, which includes root nodes, domain nodes and leaf nodes. And then an RIMS is proposed, and the resources publication and discovery mechanism are detailedly described. At last, an application prototype is developed to show the validity and the practicability of the proved theory and method.展开更多
Spectrum sharing for efficient reuse of licensed spectrum is an important concept for cognitive radio technologies.In a spectrum-sharing system(SSS),deploying the antennas in a distributed manner can offer a new spati...Spectrum sharing for efficient reuse of licensed spectrum is an important concept for cognitive radio technologies.In a spectrum-sharing system(SSS),deploying the antennas in a distributed manner can offer a new spatial dimension for the efficient reuse of licensed frequency bands.To improve the whole performance of multiple secondary users(SUs),this paper addresses the problem of coordinated multi-SU spectrum sharing in a distributed antenna-based SSS.By adopting the Hungarian method,the primal decomposition method and pricing policy,we propose a coordinated multi-user transmission scheme,so as to maximize the sum-rate of SUs.Simulation results show that the proposed method can significantly enhance the system performance,and the computational complexity is low.展开更多
This paper presents a novel spec- trum sharing design aiming at optimising the performance of a Multiuser Orthogonal Freq- uency-Division Multiplexing （MU-OFDM） Co- gnitive Radio Network （CRN） that consists of mul...This paper presents a novel spec- trum sharing design aiming at optimising the performance of a Multiuser Orthogonal Freq- uency-Division Multiplexing （MU-OFDM） Co- gnitive Radio Network （CRN） that consists of multiple secondary Transmitter-Receiver （Tx-Rx） pairs. For most MU-OFDM systems, the Exc- lusive Subchannel Assignment （ESA） is an efficient resource allocation method. Noneth- eless, it is inappropriate for the network consi- dered in this paper, because subchannels shar- ing among secondary Tx-Rx pairs can further improve the system performance. We investi- gate the Weighted Sum Rate （WSR） maximi- zation problem under the Shared Subchannel Assignment （SSA）, where each subchannel is shared by multiple secondary Tx-Rx pairs. With Lagrangian duality technique, we decompose the original resource allocation problem into sev- eral sub-problems on each subchannel and pro- pose a duality-based suhchannel sharing ap- proach. For practical realisation in the cogni- tive systems without central control entity, a distributed duality-based WSR maximization scheme is presented. Simulation results mani- fest that the proposed scheme achieves sig- nificantly better performance than ESA duality scheme.展开更多
In a peer-to-peer file-sharing system, a free-rider is a node which downloads files from its peers but does not share files to other nodes. Analyzing the free-riders’ impact on system throughputs is essential in exam...In a peer-to-peer file-sharing system, a free-rider is a node which downloads files from its peers but does not share files to other nodes. Analyzing the free-riders’ impact on system throughputs is essential in examining the performance of peer-to-peer file-sharing systems. We find that the free-riders’ impact largely depends on nodes behavior, including their online time and greed of downloading files. We extend an existing peer-to-peer system model and classify nodes according to their behavior. We focus on two peer-to-peer architectures: centralized indexing and distributed hash tables. We find that when the cooperators in a system are all greedy in downloading files, the system throughput has little room to increase while the cooperators throughput degrade badly with the increasing percent of greedy free-riders in the system. When all the cooperators are non-greedy with long average online time, the system throughput has much room to increase and the cooperators throughput degrade little with a high percent of greedy free-riders in the system. We also find that if a system can tolerate a high percent of greedy free-riders without suffering much throughput degradation, the system must contain some non-greedy cooperators that contribute great idle service capacity to the system.展开更多
基金the National High Technology Research and Development Programme of China(No2004AA104280,2006AA01Z172)
文摘A hierarchical clustered BitTorrent （CBT） system is proposed to improve the file sharing perior-mance of the BitTorrent system, in which peers are grouped into clusters in a large-scale BitTorrent-hke underlying overlay network in such a way that clusters are evenly distributed and that the peers within the cluster are relatively close to each other. A fluid model is developed to compare the performance of the proposed CBT system with the BitTorrent system, and the result shows that the CBT system can effectively improve the performance of the system. Simulation results also demonstrate that the CBT system improves the system scalabihty and efficiency while retaining the robustness and incentives of the original BitTorrent paradigm.
基金Project supported by the National Natural Science Foundation of China (No. 50335020) and the Opening Foundation of Hubei Digital Manufacturing Key Lab (No. SZ0406), China
文摘In the manufacturing grid＇s architecture, Resources Management System （RMS） is the central component responsible for disseminating resource information across the grid, accepting requests for resources, discovering and scheduling the suitable resources that match the requests for the global grid resource, and executing the requests on scheduled resources. In order to resolve the problem of resources publication and discovery in Manufacturing Grid （MGrid）, the classification of manufacturing resources is first researched after which the resources encapsulation class modes are put forward. Then, a scalable two-level resource management architecture is constructed on the model, which includes root nodes, domain nodes and leaf nodes. And then an RIMS is proposed, and the resources publication and discovery mechanism are detailedly described. At last, an application prototype is developed to show the validity and the practicability of the proved theory and method.
基金supported in part by the National Science Foundation of China for Young Scholars under grant No.61201186The National Basic Research Program undergrant No.2012AA01A502+5 种基金National Natural Science Foundation of China under grant No.61201192National S&T Major Project under grant No.2014ZX03003003-002Tsinghua-HUAWEI Joint R&D on Soft Defined Protocol StackTsinghua-HUAWEI Joint Research on 5G Air Interface TechnicalTsinghua-Qualcom joint research programIndependent innovation on Future Virtualization Platform under grant No.015Z02-3
文摘Spectrum sharing for efficient reuse of licensed spectrum is an important concept for cognitive radio technologies.In a spectrum-sharing system(SSS),deploying the antennas in a distributed manner can offer a new spatial dimension for the efficient reuse of licensed frequency bands.To improve the whole performance of multiple secondary users(SUs),this paper addresses the problem of coordinated multi-SU spectrum sharing in a distributed antenna-based SSS.By adopting the Hungarian method,the primal decomposition method and pricing policy,we propose a coordinated multi-user transmission scheme,so as to maximize the sum-rate of SUs.Simulation results show that the proposed method can significantly enhance the system performance,and the computational complexity is low.
基金ACKNOWLEDGEMENT This work was supported in part by the Na- tional Natural Science Foundation of China un- der Grants No. 60972072, No. 61340033 and the 111 Project of China under Grant No. B08038.
文摘This paper presents a novel spec- trum sharing design aiming at optimising the performance of a Multiuser Orthogonal Freq- uency-Division Multiplexing （MU-OFDM） Co- gnitive Radio Network （CRN） that consists of multiple secondary Transmitter-Receiver （Tx-Rx） pairs. For most MU-OFDM systems, the Exc- lusive Subchannel Assignment （ESA） is an efficient resource allocation method. Noneth- eless, it is inappropriate for the network consi- dered in this paper, because subchannels shar- ing among secondary Tx-Rx pairs can further improve the system performance. We investi- gate the Weighted Sum Rate （WSR） maximi- zation problem under the Shared Subchannel Assignment （SSA）, where each subchannel is shared by multiple secondary Tx-Rx pairs. With Lagrangian duality technique, we decompose the original resource allocation problem into sev- eral sub-problems on each subchannel and pro- pose a duality-based suhchannel sharing ap- proach. For practical realisation in the cogni- tive systems without central control entity, a distributed duality-based WSR maximization scheme is presented. Simulation results mani- fest that the proposed scheme achieves sig- nificantly better performance than ESA duality scheme.
基金the National High Technology Re-search and Development Program (863) of China(No. 2007AA01Z457)the Shanghai Science and Technology Development Funds (No. 07QA14033)
文摘In a peer-to-peer file-sharing system, a free-rider is a node which downloads files from its peers but does not share files to other nodes. Analyzing the free-riders’ impact on system throughputs is essential in examining the performance of peer-to-peer file-sharing systems. We find that the free-riders’ impact largely depends on nodes behavior, including their online time and greed of downloading files. We extend an existing peer-to-peer system model and classify nodes according to their behavior. We focus on two peer-to-peer architectures: centralized indexing and distributed hash tables. We find that when the cooperators in a system are all greedy in downloading files, the system throughput has little room to increase while the cooperators throughput degrade badly with the increasing percent of greedy free-riders in the system. When all the cooperators are non-greedy with long average online time, the system throughput has much room to increase and the cooperators throughput degrade little with a high percent of greedy free-riders in the system. We also find that if a system can tolerate a high percent of greedy free-riders without suffering much throughput degradation, the system must contain some non-greedy cooperators that contribute great idle service capacity to the system.
